Boltzmann Graph Ensemble Embeddings for Aptamer Libraries
Starlika Bauskar ∗, Jade Jiao ∗, Narayanan Kannan ∗, Alexander Kimm ∗, Justin M. Baker, Matthew J. Tyler, Andrea L. Bertozzi †, Anne M. Andrews ‡ · 2025
Read Boltzmann Graph Ensemble Embeddings for Aptamer Libraries by Starlika Bauskar ∗, Jade Jiao ∗, Narayanan Kannan ∗, Alexander Kimm ∗, Justin M. Baker, Matthew J. Tyler, Andrea L. Bertozzi †, Anne M. Andrews ‡ (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-21980.txt.
Description
Starlika Bauskar and seven co-authors, Jade Jiao, Narayanan Kannan, Alexander Kimm, Justin M. Baker, Matthew J. Tyler, Andrea L. Bertozzi, and Anne M. Andrews, rethink how to represent aptamers, short molecules selected for their grip on a target. Most machine learning tools freeze each molecule into its single lowest-energy shape, discarding the crowd of shapes it actually flickers between.
Their embedding instead treats a molecule as a weighted ensemble of interaction graphs, with weights set by thermodynamics, so common shapes count more and rare ones less. Tested on selection experiment data, where lab biases can make weak binders look abundant, the representation still grouped molecules sensibly and explained which substructures drove binding strength. The authors suggest it can surface overlooked low-abundance candidates worth testing in the lab.
Catalog record
| CREATOR | Starlika Bauskar ∗, Jade Jiao ∗, Narayanan Kannan ∗, Alexander Kimm ∗, Justin M. Baker, Matthew J. Tyler, Andrea L. Bertozzi †, Anne M. Andrews ‡ |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 26 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-21980